China's growth model is shifting from investment to innovation: total factor productivity's share rose from 18% to 26% in 2016–22, with the digital economy explaining about 40% of that increase; coastal provinces are innovation-driven while inland regions remain capital-intensive.
Analysis of China's Economic Growth Drivers: An Empirical Study Based on an Extended Cobb-Douglas Production Function (2010-2022)
From 2010 to 2022 China shifted toward innovation-led growth: capital-output elasticity fell (0.42→0.35), TFP's contribution rose from 18% to 26%, and the digital economy accounts for roughly 40% of the TFP gain, with coastal regions led by innovation and inland regions remaining capital-dependent.
This study uses numbers to look at what caused China's economy to grow from 2010 to 2022. It does this by building an extended Cobb-Douglas production function that includes measures for the digital economy and quality-adjusted labor force. The study shows that the capital-output elasticity dropped significantly, from 0.42 in 2010–2015 to 0.35 in 2016–2022. The contribution rate of total productivity (TFP) rose from 18% to 26%, with the digital economy making up 40% of that. Regional analysis shows that the coastal regions have been driven by innovation (± ≈ 0.31), while the inland regions still have a capital-dependent model (± ≈ 0.43). This study's method is new because it uses both migrant workers monitoring data and digital economy proxy indicators. This gives a more accurate picture of how labor quality and technology progress affect each other. The study results show that China's economy has changed from being based on investments to being based on innovations. They also have policy implications for promoting high-quality development.
Summary
Main Finding
The study finds that between 2010–2015 and 2016–2022 China’s growth dynamics shifted from capital-driven to innovation-driven. Capital-output elasticity fell from 0.42 to 0.35, while the contribution of total factor productivity (TFP) to growth rose from 18% to 26% — with the digital economy accounting for about 40% of that TFP increase. Regional decomposition shows coastal provinces are now innovation-led (innovation-related contribution ≈ 0.31), whereas inland provinces remain relatively capital-dependent (capital-related contribution ≈ 0.43).
Key Points
- Extended Cobb–Douglas production function used to incorporate:
- Quality-adjusted labor (not just headcounts).
- A digital-economy variable capturing digital capital/activities.
- Measured structural change:
- Capital elasticity fell substantially between the two subperiods (0.42 → 0.35).
- TFP’s share of growth rose from 18% to 26%; ~40% of that TFP rise is attributable to the digital economy.
- Regional heterogeneity:
- Coastal regions: growth increasingly driven by innovation/TFP (~0.31).
- Inland regions: growth still driven more by capital accumulation (~0.43).
- Methodological novelty: combines migrant-worker monitoring data with proxy indicators for the digital economy to better capture labor quality and digital-technology effects.
- Policy conclusion: evidence supports policies to accelerate innovation, digitalization, and human-capital upgrading to sustain high-quality growth.
Data & Methods
- Framework: Extended Cobb–Douglas production function with factors for:
- Capital input,
- Quality-adjusted labor input (using migrant-worker monitoring data to adjust labor quality),
- TFP augmented by a digital-economy term (proxied by multiple digital indicators).
- Estimation and decomposition:
- Growth-accounting / elasticity estimation across two periods (2010–2015 and 2016–2022).
- TFP decomposed to quantify the contribution of the digital economy to productivity growth.
- Regional decomposition to estimate coastal vs inland factor contributions.
- Data innovations:
- Use of migrant-worker monitoring data to refine labor-quality measures (skill, composition, migration dynamics).
- Use of composite proxies/indicators for regional digital-economy intensity rather than relying solely on ICT investment or internet penetration.
Implications for AI Economics
- Modeling AI-related productivity:
- The large digital-economy share of TFP growth implies that models of AI diffusion and macro productivity should explicitly include digital capital and digital-intensity measures as drivers of TFP rather than treating AI as a black-box shock.
- Measurement and data:
- Quality-adjusted labor measures (e.g., using migrant-worker and microdata) are crucial for assessing AI’s labor-market impacts; aggregate headcount measures will understate skill composition and reallocation effects.
- Proxy indicators for digital activity (platform usage, cloud adoption, AI-related service penetration) can be incorporated into growth-accounting frameworks to estimate AI’s contribution to productivity.
- Regional and distributional effects:
- Heterogeneous regional responses mean AI/digital policy should be spatially targeted: coastal/innovation hubs may capture AI gains faster, while inland regions may need capital-to-innovation transition policies (training, digital infrastructure, incentives for tech adoption).
- Policy design for AI complementarity:
- Findings support policies that combine digital infrastructure investment, human-capital upskilling, and innovation incentives to convert capital-driven growth into sustained AI-enabled productivity gains.
- Research agenda:
- Further work should adapt the study’s approach to quantify AI-specific contributions to TFP, test robustness across alternative digital proxies, and track labor-quality adjustments in response to AI adoption (displacement vs upskilling).
Assessment
Paper Typecorrelational
Evidence Strengthmedium — The paper provides systematic, quantitative decomposition of growth using novel data (migrant worker monitoring and digital-economy proxies) and standard production-function methods, which supports descriptive and associative claims about the changing drivers of growth; however, it does not establish causal identification (no plausibly exogenous variation or instrumental strategy reported) and results are sensitive to measurement choices, functional-form assumptions, and potential endogeneity between digital adoption and productivity.
Methods Rigormedium — Methodologically sound in extending the Cobb–Douglas framework, adjusting labor quality, and conducting regional decompositions; uses new data sources that plausibly improve measurement. But risks remain from omitted variables, reverse causality, aggregation bias, measurement error in digital-economy proxies, and limited robustness checks or formal identification strategies are not described.
SampleMacro- and province-level data for China covering 2010–2022, including capital and output series, a quality-adjusted labor force constructed using migrant-worker monitoring data, and proxy indicators for the digital economy; regional (coastal vs inland) decomposition and period comparisons (2010–15 vs 2016–22).
Themesproductivity innovation
GeneralizabilityFindings are specific to China and the 2010–2022 period and may not generalize to other countries or earlier/later periods., Digital-economy proxies may not capture AI-specific technologies, limiting applicability to AI-driven productivity claims., Production-function aggregation hides sectoral heterogeneity; sector-level dynamics may differ., Migrant-worker monitoring data are country-specific and may not map to labor-quality measurement in other contexts., No causal identification strategy limits external policy transferability.
Claims (8)
| Claim | Direction | Confidence | Outcome | Details |
|---|---|---|---|---|
| The capital-output elasticity dropped significantly, from 0.42 in 2010–2015 to 0.35 in 2016–2022. Firm Productivity | negative | high | capital-output elasticity (elasticity of output with respect to capital) |
from 0.42 in 2010–2015 to 0.35 in 2016–2022
0.3
|
| The contribution rate of total factor productivity (TFP) rose from 18% to 26% between the earlier and later periods. Firm Productivity | positive | high | TFP contribution rate to economic growth |
from 18% to 26%
0.3
|
| The digital economy accounted for 40% of the observed increase in TFP (i.e., made up 40% of the TFP contribution). Firm Productivity | positive | high | share of TFP contribution attributable to the digital economy |
40%
0.3
|
| Regional analysis shows coastal regions have been driven by innovation, with an estimated (innovation) coefficient of approximately 0.31. Firm Productivity | positive | high | innovation-related elasticity/coefficient in coastal regions (≈0.31) |
≈0.31
0.3
|
| Regional analysis shows inland regions remain capital-dependent, with an estimated (capital) elasticity of approximately 0.43. Firm Productivity | mixed | high | capital elasticity in inland regions (≈0.43) |
≈0.43
0.3
|
| The study's method is novel because it uses both migrant worker monitoring data and digital-economy proxy indicators, giving a more accurate picture of how labor quality and technological progress affect each other. Skill Acquisition | positive | high | measurement accuracy of labor quality and technology interaction (methodological improvement) |
0.3
|
| Overall, China's growth model shifted over 2010–2022 from being investment-driven to being innovation-driven. Firm Productivity | positive | high | structural shift in the growth model (investment-driven → innovation-driven) |
0.3
|
| The study implies policy actions to promote high-quality development based on the finding that innovation and the digital economy now play larger roles in growth. Governance And Regulation | positive | high | policy implication for promoting high-quality development |
0.05
|